Tumor hypoxia is often found in many human cancers; their presence has been implicated in radio- and chemo-resistance, a more aggressive phenotype, and is an important prognostic factor of treatment outcome. The long term goal of this proposal is to develop an image-guided therapeutic strategy, combining hypoxia imaging, radiotherapy and gene therapy, to overcome hypoxia-mediated radioresistance and improve cancer cure. The major facets of this strategy are 1) the imaging and localization of tumor hypoxia, 2) image-guided delivery of vectors that express a radiosensitizing effector gene as well as a hypoxia-induced marker gene, 3) verification of vector delivery to hypoxic cells by molecular imaging, and 4) tumor eradication by radiation. To evaluate and validate this approach, there are three Specific Aims: In Specific Aim I, we shall design and construct vectors that express both a marker gene HSV1-tk from a hypoxia-inducible promoter and an effector gene that antagonizes the repair of DNA breaks, and use them to stably transfect tumor cell lines. The degree of radiosensitization of tumor cells that constitutively or inducibly express various effector genes will be assessed in vitro and in vivo; the specific expression of the viral tk gene in hypoxic cells will be evaluated by microPET imaging of TK-mediated trapping of 124I-FIAU in tumors transplanted in nude mice. In Specific Aim II, replication defective adenovirus containing the hypoxia marker gene and the effector gene will be constructed. The efficacy of adenovirus-mediated delivery of the marker gene and the effector gene, the hypoxia inducibility of the marker tk gene, and the radiosensitizing effect of the effector gene will then be studied in vitro and in vivo. In Specific Aim III, we shall transplant rodent and human tumors into nude rats, identify and localize the hypoxic sub-volume using microPET/18F-FMISO imaging. Guided by these images, adenoviral vectors will be delivered to the hypoxic region of the tumor and microPET imaging based on TK-mediated trapping of 124I-FIAU will be used to verify the preferential delivery of the vectors to hypoxic cells. The tumor's response to radiation will then be evaluated. In pilot studies, we have shown that antisense Ku70 or a dominant negative Ku70 fragment significantly radiosensitizes hypoxic tumor cells in vitro and in vivo, and that they can be successfully delivered into tumors by adenoviral vectors. For tumor hypoxia targeting, we have implemented microPET imaging with 18F-FMISO and developed a stereotaxic template for guiding adenoviral vector injection. Also, we have shown that the hypoxia-induction of the marker tk gene can be readily detected using 124I-FIAU / microPET in vivo. The information gained from the proposed preclinical studies will serve as a guide in the design of clinical strategy to improve the outcome of cancer therapy.